JP4995686B2 - Disaster damage forecasting device, disaster damage forecasting method, and disaster damage forecasting program - Google Patents

Description

  The present invention relates to a disaster damage estimation apparatus, a disaster damage prediction method, and a disaster damage prediction program.

  Conventionally, in order to efficiently carry out restoration work in response to severe damage such as power outages, building damage, road damage, and fires caused by wide-area disasters such as earthquakes, damage assumptions after disasters have been made with high accuracy. It is requested. In particular, since the stable supply of electric power is the foundation of industry or life, it is required to realize a highly accurate damage assumption in the power distribution facility in order to support efficient restoration work.

  For this reason, for example, in Non-Patent Document 1, Non-Patent Document 2 and Patent Document 1, in order to assume damage caused by earthquakes in a sequentially updated manner (real time), sequentially updated information on earthquake motion, satellite photographs (aerial photographs) ) A technique for improving the damage estimation accuracy of structures such as power distribution facilities by utilizing information and both through image processing is disclosed.

Noroshima, Masashi Matsuoka, Shinta Sugito, Kenichi Esaki: Development of quantitative estimation method of building destruction rate by integrating ground motion information and satellite SAR image information, Proceedings of JSCE, Vol.62 No.4 pp.808-821, October 2006 Ministry of Land, Infrastructure, Transport and Tourism Comprehensive Technology Development Project, Technology Development for Rapid Disaster Prevention and Mitigation Measures Using Disaster Information and Examination of Promotion Measures, Comprehensive Report, Ministry of Land, Infrastructure, Transport and Tourism, December 2006 JP 2003-287573 A

  By the way, the above-mentioned conventional technology has a limit in the accuracy of the assumption of damage in facilities distributed over a wide range when damage is estimated using only information on seismic motion. (Aerial photograph) information is used. However, when making damage assumptions using satellite photo information, there is a problem that it is not possible to quickly make highly accurate damage assumptions after a disaster occurs.

  In other words, at present, it is difficult to obtain a satellite photograph of the affected area immediately after a disaster. Therefore, highly accurate damage estimation using satellite photograph information and earthquake motion information is possible only after several hours of disaster occurrence. Cannot be done. Therefore, at present, damage assumption using satellite photograph information and earthquake motion information is performed immediately after a disaster, and based on this, it is not possible to give an initial motion instruction or a patrol instruction in restoration work.

  In addition, here, it was described that there was a problem that high-precision damage assumption after a disaster could not be made quickly when an earthquake disaster occurred, but this is not limited to earthquakes. There were similar problems.

  Therefore, the present invention has been made to solve the above-described problems of the prior art, and by using information that can be obtained immediately after the occurrence of a disaster, damage assumption after the occurrence of the disaster can be made quickly and with high accuracy. It is an object of the present invention to provide a disaster damage estimation apparatus, a disaster damage prediction method, and a disaster damage prediction program that can be performed.

  In order to solve the above-described problems and achieve the object, the invention according to claim 1 is a disaster damage estimation apparatus that assumes damage of a predetermined structure due to a disaster that has occurred in a predetermined area, wherein the predetermined area , Disaster history information storage means for storing disaster history information, which is history information of the disaster that has occurred in the past, for each mesh preset in the predetermined area, and a support installed in the predetermined area Based on the disaster information that is the information of the disaster, from the performance value that represents the safety of the object and the regional coefficient that quantifies the regional characteristics in consideration of the ground conditions of the region where the support is installed, Damage rate function storage means for storing a damage rate function for calculating the damage occurrence rate of the support for each mesh, and the occurrence of the disaster as a node without a parent node, and the support generated by the disaster A setting network structure storage means for storing a setting network structure defined and set by a Bayesian network expressing a causal relationship between a plurality of damage events including harm and damage of the predetermined structure by a conditional probability; and the disaster Based on the immediately after disaster information that is the disaster information acquired immediately after the occurrence of the disaster and the disaster history information for each mesh stored in the disaster information storage means, the disaster information in the entire predetermined region is estimated and estimated The disaster information in the whole predetermined area, the damage information confirmed immediately after the disaster that is the damage information confirmed in the whole area obtained together with the immediately following disaster information, and the damage stored in the damage rate function storage means Based on the rate function and the setting network structure stored by the setting network structure storage means, For each mesh set in advance in the area, a damage immediately after disaster estimating means for estimating a damage occurrence rate in the predetermined structure immediately after the occurrence of the disaster, and disaster information updated after the disaster has occurred Updating the damage rate function stored in the damage rate function storage means based on the updated disaster information and the damage information of the support determined in a part of the predetermined area after the disaster has occurred; In addition to the part of the predetermined area, the assumption of the damage occurrence rate in the predetermined structure in the setting network structure stored by the setting network structure storage unit is set for each mesh preset in the predetermined area. And an assumed damage updating means for updating.

  Further, in the invention according to claim 2, in the above-mentioned invention, the damage assumption updating means acquires post-disaster confirmed damage information which is damage information confirmed in the entire predetermined area after the occurrence of the disaster. In the case, a damage event corresponding to the post-disaster confirmed damage information is added to the setting network structure stored by the setting network structure storage means, and the added setting network structure is not included in a part of the predetermined area. The assumption of the occurrence rate of damage in the predetermined structure based on the basis is updated for each mesh preset in the predetermined area.

  According to a third aspect of the present invention, in the above-described invention, a damage assumption map that displays the damage occurrence rate estimation result of the predetermined structure assumed by the damage prediction means immediately after the disaster together with the map of the predetermined area. It is further characterized by further comprising an assumed damage map creation / display means for creating and displaying on a predetermined display unit.

  According to a fourth aspect of the present invention, in the above invention, the damage assumption map creation / display means generates damage to the predetermined structure other than a part of the predetermined area assumed by the damage assumption update means. The estimated damage update result is displayed together with the damage information of the support confirmed in a part of the predetermined area and the confirmed damage information after the disaster, and the displayed damage assumption map is displayed on the predetermined display unit. It is further characterized by means for creating and displaying a damage assumption map.

  The invention according to claim 5 is the above invention, wherein the disaster is an earthquake disaster, the disaster information is earthquake motion information composed of an earthquake motion intensity distribution, and the confirmed damage information immediately after the disaster is power failure information. And the predetermined structure is the support.

  The invention according to claim 6 is the above invention, wherein the disaster is an earthquake disaster, the disaster information is earthquake motion information including a seismic intensity distribution, and the confirmed damage information immediately after the disaster is power failure information. And the predetermined structure is a building adjacent to the support.

  The invention according to claim 7 is the above invention, wherein the disaster is a typhoon disaster, the disaster information is weather information including wind speed information, and the confirmed damage information immediately after the disaster is power failure information. And the predetermined structure is the support or a building adjacent to the support.

  The invention according to claim 8 is a disaster damage estimation method that assumes damage to a predetermined structure due to a disaster that has occurred in a predetermined area, and the history information of the disaster that has occurred in the past in the predetermined area Disaster history information storage step of storing in the first storage unit every disaster mesh information that is preset in the predetermined area, and safety of the support installed in the predetermined area The support for each mesh based on the disaster information that is the information of the disaster from the performance value that represents and the regional coefficient that quantifies the regional characteristics in consideration of the ground conditions of the area where the support is installed A damage rate function storage step of storing a damage rate function for calculating a damage occurrence rate of the support in the second storage unit, and the occurrence of the disaster as a node without a parent node, and the support generated by the disaster A setting network structure for storing a setting network structure defined and set by a Bayesian network expressing a causal relationship between a plurality of damage events including harm and damage of the predetermined structure with a conditional probability in a third storage unit Based on the storage step, the immediately following disaster information that is the disaster information acquired immediately after the occurrence of the disaster, and the disaster history information for each mesh stored in the first storage unit, in the entire predetermined region Estimating disaster information, the estimated disaster information in the entire predetermined area, the confirmed damage information immediately after the disaster that is the damage information confirmed in the entire predetermined area acquired together with the immediately following disaster information, and the second Based on the damage rate function stored in the storage unit and the set network structure stored in the third storage unit, the predetermined area For each mesh set in advance, a damage assumption step immediately after a disaster for estimating a damage occurrence rate in the predetermined structure immediately after the occurrence of the disaster, and updated disaster information that is updated after the occurrence of the disaster And updating the damage rate function stored in the second storage unit based on damage information of the support determined in a part of the predetermined area after the disaster has occurred, The assumption of the damage occurrence rate in the predetermined structure in the setting network structure stored in the third storage unit is updated for each mesh preset in the predetermined area, except for a part of the predetermined area. And a damage assumption update step to be performed.

  The invention according to claim 9 is a disaster damage estimation program that causes a computer to execute a disaster damage estimation method that assumes damage to a predetermined structure due to a disaster that has occurred in a predetermined area. Disaster history information storage procedure for storing disaster history information, which is history information of the disaster that occurred in the past, in the first storage unit for each mesh preset in the predetermined area, and setting in the predetermined area The disaster information, which is the information of the disaster, is obtained from the performance value indicating the safety of the supported support and the region coefficient obtained by quantifying the regional characteristics in consideration of the ground conditions of the region where the support is installed. Based on a damage rate function storage procedure for storing a damage rate function for calculating the damage occurrence rate of the support for each mesh in a second storage unit, and a node without a parent node for the occurrence of the disaster A set network structure defined and set by a Bayesian network that expresses a causal relationship between a plurality of damage events including damage of the support and damage of the predetermined structure caused by the disaster by a conditional probability. Setting network structure storage procedure stored in the third storage unit, immediately after disaster information that is disaster information acquired immediately after the occurrence of the disaster, the disaster history information for each mesh stored in the first storage unit, Based on the above, the disaster information in the entire predetermined area is estimated, and the disaster information that is confirmed in the entire predetermined area is acquired together with the disaster information in the entire predetermined area and the immediately following disaster information. Immediately after the confirmed damage information, the damage rate function stored in the second storage unit, and the set network stored in the third storage unit. Damage prediction procedure immediately after a disaster for estimating a damage occurrence rate in the predetermined structure immediately after the occurrence of the disaster for each mesh set in advance in the predetermined area based on the structure, and the occurrence of the disaster The second storage unit stores the updated disaster information, which is updated after the disaster, and the damage information of the support determined in a part of the predetermined area after the disaster has occurred. By updating the damage rate function, the assumption of the damage occurrence rate in the predetermined structure in the setting network structure stored in the third storage unit other than a part of the predetermined region It is characterized by causing a computer to execute a damage assumption updating procedure performed by updating for each mesh preset in the area.

  According to the invention of claim 1, 8 or 9, the damage occurrence rate of the damage target structure incorporated in the Bayesian network is immediately estimated by Bayesian estimation using information immediately available immediately after the disaster occurs. In addition, the damage occurrence rate in the structure can be updated and assumed by the damage rate function updated using the support damage information in some areas determined after the disaster occurs. It is possible to make damage assumptions quickly and with high accuracy.

  In addition, according to the invention of claim 2, for example, by incorporating damage information of structures such as roads and buildings in the entire area acquired through the network after the disaster has occurred into the Bayesian network, Since the damage occurrence rate can be assumed by Bayesian estimation, it is possible to make a damage assumption after a disaster occurs with higher accuracy.

  According to the invention of claim 3, it is possible to identify an appropriate area for allocating personnel involved in recovery work based on quick and high-precision damage assumption at the time of initial recovery work immediately after a disaster. it can.

  According to the invention of claim 4, the personnel involved in the recovery work can be distributed and arranged in an appropriate region based on the more accurate damage assumption updated after the disaster occurs.

  Further, according to the invention of claim 5, the occurrence rate of damage in the support incorporated in the Bayesian network is immediately estimated by Bayesian estimation using the power failure information that can be obtained immediately after the occurrence of the earthquake disaster, and the occurrence of the disaster Support damage after the occurrence of an earthquake disaster can be assumed by updating the damage occurrence rate of support in an unconfirmed area with the damage rate function updated using support damage information for a certain area that was confirmed later. It is possible to make damage assumptions quickly and with high accuracy.

  According to the invention of claim 6, Bayesian estimation of the damage occurrence rate of the building that may cause damage to the support incorporated in the Bayesian network using the power outage information immediately available immediately after the occurrence of the earthquake disaster. As a result, the damage occurrence rate of support in unidentified areas was updated using the damage rate function updated using support damage information for certain areas determined after the occurrence of an earthquake disaster. Since building damage can be updated and assumed based on the damage occurrence rate of the support, building damage can be predicted quickly and accurately after an earthquake disaster occurs.

  According to the invention of claim 7, the damage occurrence rate in the support incorporated in the Bayesian network is immediately assumed by Bayesian estimation using the power failure information that is immediately available immediately after the occurrence of the typhoon disaster. Support rate after typhoon disaster has occurred because it can be assumed that the damage occurrence rate of support in unidentified areas can be updated with the damage rate function updated using support damage information for some areas later determined. It is possible to make damage assumptions quickly and with high accuracy. Or, immediately after the occurrence of a typhoon disaster, we will immediately estimate the damage occurrence rate of buildings that may cause damage to the support incorporated in the Bayesian network using the power outage information that is immediately available. Update the damage occurrence rate of support in unconfirmed areas with the updated damage rate function using support damage information for some areas determined after the occurrence, and based on the updated support damage occurrence rate Since building damage can be updated and assumed, building damage prediction after a typhoon disaster can be made quickly and with high accuracy.

  Exemplary embodiments of a disaster damage prediction apparatus, a disaster damage prediction method, and a disaster damage prediction program according to the present invention will be described below in detail with reference to the accompanying drawings. In the following, an earthquake disaster damage estimation apparatus that performs damage estimation when an earthquake disaster occurs will be described.

  The earthquake disaster damage forecasting device in this example is based on the assumption that the damage of the support due to the earthquake disaster that occurred in the target area of the damage forecast is as an overview, and the damage forecast after the earthquake disaster occurs can be performed quickly and with high accuracy. The main feature is that it becomes possible. Hereinafter, this main feature will be described with reference to FIG. 1 which is a diagram for explaining the configuration of the earthquake disaster damage estimation apparatus.

  As shown in FIG. 1, the earthquake disaster assumption apparatus 10 in this embodiment includes an input unit 11, an output unit 12, an input / output control I / F unit 13, a storage unit 14, and a processing unit 15. The

  The input unit 11 inputs various kinds of information, and particularly those closely related to the present invention include “earthquake motion information” composed of seismic motion intensity distribution, “blackout information” as confirmed damage information immediately after a disaster, and occurrence of an earthquake. “Seismic ground motion information” updated later, support damage information based on patrol, etc. are received and input via a network such as the Internet or an intranet, for example, and various information stored in the storage unit 14 to be described later Are also received and input from the administrator of the earthquake disaster assumption apparatus 10.

  The output unit 12 outputs various types of information and is configured to include a monitor and a speaker. Particularly, as closely related to the present invention, an earthquake damage assumption map created by an earthquake damage assumption map creation unit 15c described later is used. The information is displayed on the monitor screen together with the support damage information confirmed by patrol and the post-disaster confirmed damage information such as building damage confirmed throughout the target area.

  The input / output control I / F unit 13 controls data transfer between the input unit 11 and the output unit 12, the storage unit 14, and the processing unit 15.

  The storage unit 14 stores data used for various types of processing by the processing unit 15 and various types of processing results by the processing unit 15, and as particularly closely related to the present invention, as shown in FIG. An earthquake history storage unit 14a, a damage rate function storage unit 14b, and a setting network structure storage unit 14c are provided.

  As shown in FIG. 2, the target area map / earthquake history storage unit 14a applies a map of the target area for damage assumption and earthquake disaster history information that is history information of earthquake disasters that occurred in the past in the target area. It memorize | stores for every mesh preset in the area. Here, FIG. 2 is a diagram for explaining the target area map / earthquake history storage unit.

  For example, as shown in FIG. 2A, the target area map / earthquake history storage unit 14a is preliminarily placed in the target area together with a map of the target area for damage assumption (see the area surrounded by a dotted line in the figure). For example, the position of the mesh divided into “500 m × 500 m” sections is stored. Here, although not shown in the figure, the target area map / earthquake history storage unit 14a is provided with support information, building layout information, and lifeline layout information such as roads, water pipes, and gas pipes in the target area. I also remember. In addition, in this embodiment, the support means specifically a wire for supplying electric power, a common wire, and a power distribution column for connecting various power distribution devices.

  Further, as shown in FIGS. 2B and 2C, the target area map / earthquake history storage unit 14a stores the seismic intensity distribution in the past earthquake disaster for each mesh. Here, in (B) and (C) of FIG. 2, the shading on the mesh represents the magnitude of the seismic intensity. As the earthquake motion intensity, for example, Gal, which is a unit of acceleration used to express the intensity of earthquake shaking, is used. The target area map / earthquake history storage unit 14a may also store the epicenter position for each earthquake disaster that has occurred in the past.

  Returning to FIG. 1, the damage rate function storage unit 14b is a region that takes into consideration the performance value representing the safety of the support installed in the target area, the ground conditions of the area where the support is installed, and the like. A damage rate function for calculating the damage occurrence rate of the support for each mesh is stored based on the earthquake disaster information from the regional coefficient whose characteristics are quantified. Hereinafter, the damage rate function stored in the damage rate function storage unit 14b will be described with reference to FIGS. 3 is a diagram for explaining the damage rate function, FIG. 4 is a diagram for explaining the area coefficient in the damage rate function, and FIGS. 5 and 6 explain the performance value in the damage rate function. It is a figure for doing.

  Here, according to the empirical probability distribution in the occurrence of past earthquake disasters, even if the seismic intensity is the same, the performance value (z) of the support in which the damage actually occurred is widely distributed. Whether or not damage has occurred can be determined not only by the performance value (z), but it is also necessary to consider the ground conditions where the support is installed and the effects of damage to surrounding facilities such as flying objects, trees and buildings collapsed. For this reason, the probability of damage (p) in which damage such as breakage or inclination of the support due to earthquake motion occurs, and the performance value (z) defined by the “base fracture moment / maximum generated bending moment” of the support, The relationship is desirably modeled by a damage rate function (p (z)) that can be described by a curve shown in FIG. 3A so that uncertain factors can be considered probabilistically.

Therefore, in the present invention, the damage probability (p _i (z)) in consideration of the ground conditions is defined by the mathematical formula shown in FIG. In FIG. 3B, “p _i (z)” is the damage probability of the support that is installed in the regional characteristic “i” reflecting the ground condition and has the performance value “z”, and “F _i ( z) "is a reference hazard function for performance value to" z "and parameter" r _i "is the area coefficient of regional characteristics" i ". Here, “F _i (z)” is a distribution function obtained by integrating the probability distribution function with the performance value “z” as shown in FIG.

Moreover, regional coefficient (r _i) is the ordinary soil or soft soil, such as ground conditions, the addition of urban and mountain areas regional characteristics such as, in the mechanical model taking into account the variation in the presence or absence of damage that can not be explained For example, as shown in Fig. 4, damage to the support breakage depends on the ground motion intensity range for each ground condition with or without liquefaction potential, as shown in Fig. 4. It is set based on experience values for each mode and damage mode of support slope.

  That is, as shown in FIG. 4, when an earthquake with a seismic intensity (200 to 249 gal) occurs in an area where the liquefaction possibility index (PL) is greater than “15” and the liquefaction is likely to occur. In the damage mode of support breakage, “r13” is set, and in the damage mode of support tilt, “r14” is set. In addition, when an earthquake with seismic intensity (200 to 249 gal) occurs in an area where liquefaction possibility index (PL) is greater than “5” and less than or equal to “15” and the possibility of occurrence of liquefaction is moderate. Is set to “r23” in the damage mode of support breakage, and is set to “r24” in the damage mode of support tilt. If an earthquake with seismic intensity (200 to 249 gal) occurs in a non-liquefied area where liquefaction is unlikely to occur, “r33” is set in the damage mode of support breakage. In the damage mode of the object inclination, “r34” is set. Note that the above “r13” is represented by a numerical value such as “0.00008”, for example. Also, in Fig. 4, the seismic intensity is classified into four ranges, and regional factors are set according to the degree of possibility of liquefaction classified into three for each of the four classified seismic intensity ranges. However, the present invention is not limited to this, and the range of seismic intensity and the degree of possibility of liquefaction may be classified into an arbitrary range and an arbitrary level. The coefficient is set according to these arbitrarily set classifications.

Further, as shown in FIG. 5A, the performance value (z) is a distribution obtained from “maximum ground motion acceleration” as “static horizontal seismic force (load (kgf) converted to top concentrated load of support)”. This is obtained by multiplying “Z ₀ ”, which is a safety factor against destruction of a single utility pole (support), by a coefficient composed of “k1 to k4.” Here, as shown in FIG. "Is a coefficient that takes into account the influence of overhead wires, co-wires, and various devices, and is calculated based on the masses of these wires, co-wires, and various devices and the height from the ground (Fig. 5). In addition, as shown in FIG. 5B, “k2” is a response magnification with respect to the maximum acceleration of the natural period of the distribution column, and “k3” is a coupled effect. “K4” is a correction coefficient that takes into account the influence of adjacent columns.

Further, “Z ₀ ” that is the safety factor of the support is calculated based on, for example, “Electrical Technical Standards Investigation Committee, Distribution Regulations (Low Voltage and High Voltage) JEAC 7001-1992, NEC Corporation”. That is, “Z ₀ ”, which is the safety factor of the support, is expressed by respective formulas when the support is not rooted and when the support is rooted as shown in FIG. Is calculated according to In the mathematical formula shown in FIG. 6A, “D ₀ ” is the diameter (unit: m) at the base of the support, and “t” is the penetration depth (unit: m) of the support. Yes, “H” is the height (unit: m) on the ground surface of the concentrated weight point, and “P” is a load (unit: kgf) converted to the top concentrated weight of the support.

Here, “t ₀ ” is the depth from the ground surface to the center of rotation of the support when the support is not rooted, and is calculated by the mathematical formula shown in FIG. “T ₀ ′” is the depth from the ground surface to the center of rotation of the support when the support is rooted, and the depth from the ground surface to the center of the root (unit: m) In the case of “t _c ”, it is calculated by the mathematical formula shown in FIG. In addition, "n" used by the numerical formula shown in FIG. 6C is a ratio obtained from the root area (A) and the area of the support base (A ₁ ), and the root area (A) The mathematical formula for calculating the area (A ₁ ) of the support base is also shown in FIG. Further, “Q” and “J” used in the mathematical formula shown in FIG. 6A are calculated by the mathematical formula shown in FIG.

Furthermore, in the mathematical formula shown in FIG. 6A, “K” is a constant representing a ground condition determined from a dynamic experiment, is a soil coefficient, and is 4.0 × 10 ^{6 in} the case of ordinary soil [A]. (kgf / m ^4), in the case of ordinary soil ^{[B], 3.0 × 10 6} (kgf / m 4), in the case of soft soil ^{[C], 2.0 × 10 6} (kgf / m 4), soft soil [ In the case of D], 0.8 × 10 ⁶ (kgf / m ⁴ ) is given. In the mathematical formula shown in FIG. 6A, “P” is calculated by the mathematical formula shown in FIG. In the equation shown in FIG. 6E, “E” is the maximum ground motion acceleration, “H” is the ground height (unit: m) of the support, and “W ₁ ” is the transformer. , “W ₂ ” is the weight of the ground part of the support (unit: kg), “w” is the weight of the crossover line (unit: kg / m), “L ₁ ” is the ground clearance (unit: m) of the transformer mounting part, “l ₂ ” is the ground height of the center of gravity of the support ground part (unit: m), and “h” is the interference The ground height of the line (unit: m), and “s” is the sum of the diameters on both sides divided by 2 (unit: m).

Thus, based on the mathematical formulas described with reference to FIGS. 3 to 6, the damage probability “p _i (z) of the support installed at the regional characteristic“ i ”reflecting the ground condition and having the performance value“ z ”. Is defined, and the damage rate function storage unit 14b illustrated in FIG. 1 stores “p _i (z)”.

  Returning to FIG. 1, the setting network structure storage unit 14c sets the occurrence of an earthquake disaster (earthquake motion) as a node without a parent node, and sets a causal relationship between a plurality of damage events including damage to a support caused by the earthquake disaster. The setting network structure defined and set by the Bayesian network expressed by the attached probability is stored. For example, the setting network structure storage unit 14c stores a setting network structure as shown in FIG. Here, FIG. 7 is a diagram for explaining the setting network structure storage unit.

  As shown in FIG. 7A, the setting network structure storage unit 14c uses the occurrence of “earthquake motion” in consideration of ground conditions such as liquefaction as a parent node, and “support damage” caused by the earthquake motion. And the causal relationship with “power failure” caused by “support damage” is stored as a network. At that time, the conditional probability function in the causal relationship of “support damage” due to “earthquake motion” is stored as “L2”, and the conditional probability function in the causal relationship of “blackout” due to “support damage” is stored as “L3”. . Then, it is stored assuming that “support damage” is assumed for each mesh set in the target area.

  In addition to the setting network structure shown in FIG. 7A, the setting network structure storage unit 14c is connected to the causal relationship of “support damage” due to “building damage” as shown in FIG. A setting network structure in which the conditional probability function is “L1” is also stored. Here, in the setting network structure shown in FIG. 7B, when the building damage information in the entire target area is input as the post-disaster confirmed damage information from the input unit 11, the processing by the damage assumption update unit 15b described later is performed. Used for.

  In addition to the setting network structure shown in FIG. 7B, the setting network structure storage unit 14c is connected to the causal relationship of “support damage” caused by “road damage” as shown in FIG. The conditional probability function is “L4”, the conditional probability function in the “cause of damage to support” due to “fire” is “L5”, and the condition in the causal relationship of “support damage” due to “other infrastructure damage” A setting network structure having a probability function “L6” is also stored. Here, in the setting network structure shown in FIG. 7C, when information on building damage, road damage, fire, and other infrastructure damage in the entire target area is inputted from the input unit 11 as post-disaster confirmed damage information, It is used for processing by a damage assumption update unit 15b described later.

  Although not shown in the figure, information such as building damage, road damage, fire, and other infrastructure damage is sequentially entered as “damage events” in FIG. The setting network structure storage unit 14c stores setting network structures of all patterns.

  The processing unit 15 executes various processes based on the data stored in the storage unit 14, and particularly as closely related to the present invention, as shown in FIG. The update part 15b and the earthquake damage assumption map preparation part 15c are provided.

  Immediately after the disaster, the disaster assumption unit 15a includes the immediately after disaster information (earthquake motion intensity distribution) acquired immediately after the occurrence of the earthquake disaster, and the earthquake disaster history information (earthquake motion intensity distribution) for each mesh stored in the target area map / earthquake history storage unit 14a. ) To estimate the seismic intensity distribution in the entire target area. For example, when the immediately after disaster information is input from the input unit 11 as the Meteorological Agency information, the target area map / earthquake history storage unit 14a such as (B) or (C) in FIG. With reference to the earthquake disaster history information, the seismic intensity distribution in the entire target area is estimated for each mesh using the distance attenuation formula. Alternatively, when the immediately after disaster information is input from the input unit 11 as the Meteorological Agency information, the target area map / earthquake history storage unit 14a such as (B) or (C) in FIG. With reference to the earthquake disaster history information, the seismic intensity distribution is estimated for each mesh on the basis of the data approximated to the disaster information immediately after.

  Thus, as shown in FIG. 8A, the disaster immediately after disaster assumption unit 15a estimates the seismic intensity distribution in the entire target area for each mesh. Here, FIG. 8 is a figure for demonstrating the damage assumption part immediately after a disaster.

  Then, the damage immediately after disaster assumption unit 15a calculates the earthquake motion intensity distribution in the entire target area estimated, the power outage information determined in the entire target area acquired together with the immediately following disaster information, and the damage rate stored in the damage rate function storage unit 14b. Based on the function and the setting network structure stored in the setting network structure storage unit 14c, the damage occurrence rate of the support immediately after the occurrence of the earthquake disaster is estimated for each mesh or for each support in the mesh. Do.

  For example, as shown in FIG. 8B, when the power failure confirmation information confirmed in the entire target area is input via the input unit 11, the damage immediately after disaster assumption unit 15 a receives support from this information. Damage is inversely estimated by the probability propagation method in the Bayesian network, and the damage rate of the support is estimated for each mesh or for each support in the mesh.

  When earthquake motion occurs, the conditional probability table in the setting network shown in FIG. 7A is as shown in FIG. Here, FIG. 9 is a diagram for explaining the conditional probability table used in the damage assumption unit immediately after a disaster. In FIG. 9, the probability variable for “earthquake motion” is “X1”, the probability variable for “support damage” is “X2”, the probability variable for “power failure” is “X3”, and the occurrence of an event “T ", A non-occurrence where no event occurs is denoted as" F ".

  As shown in FIG. 9, when earthquake motion occurs, the probability of occurrence of earthquake “P (X1) = T” is “1” and “P (X1) = F” is “0”. Then, as shown in FIG. 8A, the earthquake motion intensity distribution in the entire target area estimated for each mesh is input, and the damage assumption unit 15a immediately after the disaster stores the damage stored in the damage rate function storage unit 14b. Using the rate function, the support damage probability “A” based on the regional characteristics for each mesh is calculated. Then, when earthquake motion occurs, the probability “P (X2) = T” of occurrence of support damage is “A”, and the probability “P (X2) = F” of no support damage occurs is “1-A”. It becomes. If there is no earthquake motion, the probability of occurrence of support damage “P (X2) = T” is “0”, and the probability of occurrence of support damage “P (X2) = F” is “1”. . When “support damage” occurs (X2 = T), the probability that a power failure does not occur (P (X3 = F)) is “0”, and the probability that a power failure occurs (P (X3 = T)). ) Becomes “1”. When the total number of supports belonging to the feeder in the target mesh is “C”, if “support damage” does not occur (X2 = F), the probability of no power failure (P (X3 = F)) is , “(1−A) to the (C−1) th power”, and the probability of power failure (P (X3 = T)) is obtained by subtracting (1−A) to the (C−1) th power. Value ".

  Then, as shown in FIG. 8B, when the power failure confirmation information confirmed in the entire target area is input via the input unit 11, the occurrence state (T or F) of power failure for each mesh is confirmed. Based on this information, the immediately after disaster damage estimation unit 15a performs the inverse estimation using the conditional probability function “L3”, specifically, the probability propagation method for each mesh or support within the mesh. Estimate support damage incidence for each item.

  Here, in the following, an estimation method using a Bayesian network performed by the damage estimation unit 15a immediately after a disaster will be described with reference to FIGS. FIG. 10 and FIG. 11 are diagrams for explaining an estimation method using a Bayesian network that is used in a damage assumption unit immediately after a disaster.

  By modeling the problem using a Bayesian network, the target including uncertainty (in this example, the situation of support damage) is represented on the computer as a set of random variables and a conditional probability between them. Calculation operations are possible with. At this time, it is a feature of the Bayesian network that it is possible to evaluate the probability distribution of the probability that each state that an arbitrary variable can take without distinguishing between an explanatory variable and an objective variable. Here, the joint probability distribution of the Bayesian network can be expressed by the mathematical formula shown in FIG.

  Next, the basics of probability inference will be described. In probability inference, a model is defined by a conditional probability distribution group assigned to each node. That is, a conditional probability table is defined (see, for example, FIG. 9). Generally describing a conditional probability table, k discrete states (where j is a vector pattern composed of various parent node groups) where pa (Xi) = y (y is a parent node group) The conditional probability distribution of the variable Xi having 1,..., K) is expressed as shown in FIG. However, the sum of these is “1”. For calculation, the conditional probability table is rewritten for each event to be a causal basic table.

And the probabilistic inference method by Bayesian network gives prior probability distribution to nodes that have neither parent node nor observation value (given by existing estimation model with seismic motion as input), and the value of observed variable “e” The posterior probability distribution P (X / e) of the variable “X” to be estimated is set in the node (in this embodiment, the variable “X2” shown in FIG. 9 is the estimation target). That is, when the value of "e" is observed as a state variable "X _e", the variable "X _e ', as shown in the first column of FIG. 10 (C), is given. This corresponds to the earthquake occurrence probability “P (X1) = T” shown in FIG. 9 being “1” and “P (X1) = F” being “0”. This “X _e ” is fixed, and a probability value for each state for other variables is obtained.

  Here, as shown in FIG. 10 (C), when a model composed of the causal relationship of three events of variables “X1 to X3” is considered as an example, P (X2 = 1) (for example, damage to the support occurs. The conditional probability) can be expressed by using a marginalization function based on the enumeration method as shown in FIG. FIG. 10C also shows a marginalization function that is a generalization of this.

Then, in order for the mathematical formula shown in FIG. 11A to approximately obtain P (X2 = 1) shown in FIG. 10C, the probability propagation from the observed information (local calculation between variables) Is a formula of a probability propagation method in which the probability distribution of each variable is updated. Note that “λ” and “π” shown in FIG. 11A are conditional probability functions between events shown in the model shown in FIG. In addition, “α”, “P (X2 | e)”, “P (X2 | e ⁺ )”, and “P (e ⁻ | X2)” in the equation shown in FIG. (B) to (E).

  In this way, the damage immediately after the disaster estimation unit 15a determines the damage to the support from the power failure confirmation information confirmed in the entire target area shown in FIG. 8B, as shown in FIG. 11A. Inverse estimation is performed by the probability propagation method in, and the damage rate of the support is estimated for each mesh or for each support in the mesh.

  Returning to FIG. 1, the earthquake damage assumption map creation unit 15 c displays the damage occurrence rate assumption result of the support assumed by the damage assumption unit 15 a immediately after the disaster, for example, for each mesh as shown in FIG. Then, an earthquake damage assumption map is displayed in which the shades corresponding to the magnitude of the damage occurrence rate are displayed together with the map of the target area, and further displayed on the monitor of the output unit 12. Here, together with the earthquake damage assumption map, road maps and lifeline information such as water pipes and gas pipes may be separately displayed in a layer format on a monitor.

  After the occurrence of an earthquake disaster, the damage assumption update unit 15b, for example, after updated earthquake disaster information updated from a strong earthquake network (K-net) operated by the National Institute for Disaster Prevention Science and Technology, and after an earthquake disaster has occurred The damage rate function stored in the damage rate function storage unit 14b is updated based on the damage information of the support determined in a part of the target area (patent end area) by patrol by the dispatched recovery worker. Thus, in the unpatrolled area other than the patrol end area, the assumption of the damage occurrence rate in the support in the setting network structure stored in the setting network structure storage unit 14c is updated for each mesh. Hereinafter, the damage assumption update unit 15b will be described with reference to FIGS. 12 and 13 are diagrams for explaining the damage assumption update unit.

  For example, as shown in FIG. 12A, the damage assumption update unit 15b acquires support damage information (patrolling information) determined in a patrol end area by patrol by a dispatched recovery worker. It is assumed that information is input via the input unit 11. Here, in FIG. 12A, it is assumed that a circle is a place of a damaged support where breakage or the like is actually confirmed. Although not shown in the figure, the earthquake immediate damage assumption unit 15a calculates the seismic intensity distribution in the entire target area shown in FIG. 8A based on the updated earthquake disaster information input via the input unit 11. It is assumed that it has been updated and estimated. In this situation, the damage assumption update unit 15b updates the damage rate function stored in the damage rate function storage unit 14b, so that the damage occurrence rate assumption of the support can be performed for each mesh in the unpatented area. Update for each support in the mesh.

  In this embodiment, facility damage such as support is modeled by a Bernie trial sequence, and the following three preconditions are set. The first is that there are two possible outcomes for equipment damage: “damaged” or “no damage”, and the second is that the probability of the target damage occurring is the same attribute. The equipment having (performance value) is assumed to be constant throughout the equipment, and the third is that each equipment damage having the same attribute is stochastically independent. Here, a method of updating the damage rate function by the damage assumption updating unit 15b based on the above three preconditions will be described with reference to FIG.

  (A) and (B) of FIG. 13 explain the names of variables used in (C), (D) and (E) of FIG.

Based on the seismic intensity information obtained immediately after the earthquake and the damage rate function stored in the damage rate function storage unit 14b (see FIG. 3B), the damage probability “p _i (z)” is immediately estimated, It is assumed that an average value and a standard deviation (see μ and σ shown in FIG. 13B) are given as indexes of variation in the estimated value of damage probability. At this time, a beta distribution is assumed as a conjugate prior distribution of the damage probability “p _i (z)”. At this time, using the Bayesian update process, the average value and standard deviation of damage probability (see μ ′ and σ ′ shown in FIG. 13 (B)) after reflecting the confirmed information of the support damage location is Are represented by mathematical formulas shown in (C) and (D) of FIG.
However, the total number of facilities with a performance value z having a local characteristic i that has been patroled and determined whether there is damage, and the number of damaged facilities within the patrol range are expressed by the equation shown in FIG. The

It should be noted that the average value of the posterior damage probability of the performance value “z” with the calculated regional characteristic “i” is applied to the unpatented area as an updated damage probability by the formula shown in FIG. . In addition, the prior distribution of damage probability based on immediate estimation is interpreted as equivalent to obtaining information that damage at n _i ⁰ (z) 'per M _i ⁰ (z)' was confirmed. The Further, the equation shown in FIG. 13C means that it is a weighted average of the average (prior average) of the sample average of the binomial distribution and the beta distribution which is the prior distribution.

  Returning to FIG. 1, the earthquake damage assumption map creating unit 15 c shows the assumed damage occurrence rate update result of the support in the unpatented area assumed by the damage assumption updating unit 15 b, for example, as shown in FIG. Then, an earthquake damage assumption map displayed together with the damage information of the support determined in the inspection end area is created and further displayed on the monitor of the output unit 12. Here, together with the earthquake damage assumption map, road maps and lifeline information such as water pipes and gas pipes may be separately displayed in a layer format on a monitor.

  Furthermore, when the post-disaster confirmed damage information (for example, building damage), which is damage information determined in the entire target area after the occurrence of the earthquake disaster, is acquired, the damage assumption update unit 15b acquires the setting network structure storage unit 14c. Is stored in the setting network structure shown in FIG. 7A, and “building damage” is added to the assumption of the damage occurrence rate in the support based on the setting network structure shown in FIG. Alternatively, update for each support in the mesh. That is, when the building damage determined in the entire target area represented by the square mark in FIG. 12C is input via the input unit 11, the damage assumption update unit 15b, based on this information, The assumption of damage occurrence rate in the support is updated for each mesh or for each support in the mesh by inverse estimation using the conditional probability function “L1” shown in FIG. . Returning to FIG. 1, the earthquake damage assumption map creation unit 15 c displays the assumed damage occurrence rate update result in the unpatented area assumed by the damage assumption update unit 15 b, for example, in (C) of FIG. 12. An earthquake damage assumption map as shown is created and displayed on the monitor of the output unit 12.

  Next, the process of the earthquake disaster damage assumption apparatus 10 in a present Example is demonstrated using FIG. FIG. 14 is a diagram for explaining processing of the earthquake disaster damage estimation apparatus in the present embodiment.

  As shown in FIG. 14, the earthquake disaster damage assumption apparatus 10 in the present embodiment, when the earthquake motion information and the power failure information immediately after the occurrence of the earthquake are input via the input unit 11 (Yes in step S1401), the damage assumption immediately after the disaster. The unit 15a estimates the seismic intensity distribution in the entire target area, and makes a support damage assumption immediately after the occurrence of the earthquake disaster for each mesh (or for each support in the mesh) (step S1402). At this time, the earthquake damage assumption map creation unit 15c uses the earthquake damage assumption map as shown in FIG. 8C, for example, as a result of the assumed damage occurrence rate of the support assumed by the damage assumption unit 15a immediately after the disaster. It is created and further displayed on a predetermined display unit on the monitor of the output unit 12.

  And the damage assumption update part 15b will update and input the patrol information after the occurrence of an earthquake (damage support information in a patrol end area), earthquake motion information, or power outage information via the input part 11 (step S1403). (Yes), by updating the damage rate function stored in the damage rate function storage unit 14b, the support damage assumption is made for each mesh (or for each support in the mesh) in the unpatrolled area other than the patrol end area. ), Updated (step S1404, sequential update estimation after earthquake disaster). Here, the damage assumption updating unit 15b also performs post-earthquake damage information (for example, building damage) after the earthquake disaster, even if the post-disaster confirmed damage information (for example, building damage) that is confirmed in the entire target area after the occurrence of the earthquake disaster is acquired. The support damage assumption is updated for each mesh (or for each support in the mesh) by successive update estimation. Further, at this time, the earthquake damage assumption map creating unit 15c displays the assumed damage occurrence rate update result of the support in the unpatented area assumed by the damage assumption updating unit 15b, for example, as shown in FIG. Then, the earthquake damage assumption map displayed together with the damage information of the support determined in the patrol end area is created, and further displayed on a predetermined display unit on the monitor of the output unit 12.

  Here, when the damage assumption update unit 15b does not determine the damage in all target areas (No in step S1405), the conditional probability function (for example, FIG. 7) in the setting network structure stored in the setting network structure storage unit 14c is stored. (L2 etc. shown in (A)) is updated (step S1406), and the process returns to step S1403 and waits until update information is input.

  On the other hand, the damage assumption update unit 15b determines the damage status map indicating the damage rate of the support and the damage location when the damage in all the target areas is confirmed (Yes in Step S1405), (Step S1407) The conditional probability function (for example, L2 shown in FIG. 7A) in the setting network structure stored in the setting network structure storage unit 14c is updated (step S1408), and the process ends.

  As described above, in the present embodiment, the immediately after disaster damage estimation unit 15a performs the immediately following disaster information acquired immediately after the occurrence of the earthquake disaster and the earthquake disaster for each mesh stored in the target area map / earthquake history storage unit 14a. Based on the history information, the seismic intensity distribution in the entire target area is estimated, the estimated seismic intensity distribution in the entire target area, the power outage information confirmed in the entire target area acquired together with the disaster information immediately afterward, and the damage rate function Based on the damage rate function stored in the storage unit 14b and the set network structure stored in the set network structure storage unit 14c, the damage occurrence rate of the support immediately after the occurrence of the earthquake disaster is estimated for each mesh. Then, the damage assumption updating unit 15b performs damage based on the updated earthquake disaster information updated after the occurrence of the earthquake disaster and the support damage information determined in the inspection end area after the occurrence of the earthquake disaster. By updating the damage rate function stored in the rate function storage unit 14b, the assumption of the damage occurrence rate in the support in the setting network structure stored in the setting network structure storage unit 14c in an unpatrolled area other than the inspection end area, Update every mesh. Further, the damage assumption update unit 15b, when post-disaster confirmed damage information (for example, building damage), which is damage information determined in the entire target area after the occurrence of the earthquake disaster, is acquired, is set network structure storage unit 14c Assuming the damage occurrence rate in the support based on the set network structure to which “building damage” stored in is added is updated for each mesh. And the earthquake damage assumption map preparation part 15c shows the damage occurrence rate assumption result of the support assumed by the damage assumption part 15a and the damage assumption update part 15b immediately after a disaster, for example, as shown in FIG. The earthquake damage assumption map displayed together with the damage information of the support determined in the patrol end area is created and further displayed on the monitor of the output unit 12.

  For this reason, the damage occurrence rate of the support incorporated in the Bayesian network is immediately estimated by Bayesian estimation using the power outage information that is immediately available immediately after the occurrence of the earthquake disaster. By using the damage rate function updated using the support damage information of the local area, the damage occurrence rate of the support in the unidentified area can be updated and assumed. It becomes possible to carry out with high precision. In addition, for example, by incorporating damage information such as buildings in the entire area acquired through the network after an earthquake disaster into the Bayesian network, the damage occurrence rate of the damage target structure can be estimated by Bayesian estimation. This makes it possible to make damage assumptions after a disaster with higher accuracy. In addition, at the time of the initial operation of recovery work immediately after the occurrence of an earthquake disaster, it is possible to identify an appropriate area for assigning personnel involved in the recovery work based on quick and highly accurate damage assumptions. In addition, based on the more accurate damage assumption updated after the occurrence of an earthquake disaster, the personnel involved in the restoration work can be distributed and arranged in appropriate areas.

  In addition, although the present Example demonstrated the case where support damage was assumed by a Bayesian network, this invention is not limited to this, For example, in the case of assuming the building damage which causes support damage. There may be. For example, it is conceivable to use a setting network structure in which a Bayesian network is defined and set as shown in FIG. 15, which is a diagram for explaining another embodiment. For example, in a setting network structure as shown in FIG. 15A, after an earthquake occurs, power outage information, support damage information determined in the inspection area and support damage assumption information assumed in the uninspected area Based on this, building damage can be assumed by inverse estimation using a conditional probability function (L1). Further, in the setting network structure as shown in FIG. 15A, after the occurrence of an earthquake, a conditional probability function (based on power failure information and support damage information in the entire area under the target area determined in the inspection area) Building damage can be assumed by inverse estimation using L1).

  For this reason, we immediately estimate the damage occurrence rate of buildings that may cause damage to the support incorporated in the Bayesian network using the power outage information that can be obtained immediately after an earthquake disaster. At the same time, the damage occurrence rate of the support in the unconfirmed area is updated by the damage rate function updated using the support damage information in some areas determined after the occurrence of the earthquake disaster. Since building damage can be updated and assumed based on the rate, building damage after an earthquake disaster can be estimated quickly and with high accuracy.

  Further, as shown in FIG. 15B, a load dropout amount, which is a dropout amount of power demand before and after an earthquake, defines a causal relationship as an event caused by building damage, and this causal relationship is a conditional probability. Let it be represented by (L7). The load dropout amount is information that can be obtained immediately after a disaster at the power company along with the power outage information, so after the occurrence of the earthquake, the power outage information, the support damage information determined in the inspection area, and the load dropout amount Based on this, building damage can be assumed by inverse estimation using conditional probability functions (L1, L7).

  Further, in this embodiment, a case has been described in which damage is assumed for a structure when an earthquake disaster occurs as a disaster. However, the present invention is not limited to this, and when a typhoon disaster occurs as a disaster. In addition, it may be a case of assuming damage to the structure. In this case, it is necessary to use a damage rate function based on wind power and regional coefficients. As a result, the damage occurrence rate of the support incorporated in the Bayesian network is immediately estimated by Bayesian estimation using the power outage information that can be obtained immediately after the typhoon disaster. The damage occurrence function updated using support damage information can be used to update and assume support damage occurrence rates in unconfirmed areas, so support damage assumptions after typhoon disasters can be made quickly and accurately. It becomes possible to do. Or, immediately after the occurrence of a typhoon disaster, we will immediately estimate the damage occurrence rate of buildings that may cause damage to the support incorporated in the Bayesian network using the power outage information that is immediately available. Update the damage occurrence rate of support in unconfirmed areas with the updated damage rate function using support damage information for some areas determined after the occurrence, and based on the updated support damage occurrence rate Since building damage can be updated and assumed, building damage prediction after a typhoon disaster can be made quickly and with high accuracy.

  In addition, among the processes described in this embodiment, all or part of the processes described as being performed automatically can be performed manually, or the processes described as being performed manually can be performed. All or a part can be automatically performed by a known method. In addition, the processing procedure, control procedure, specific name, and information including various data and parameters shown in the above-described document and drawings can be arbitrarily changed unless otherwise specified.

  Further, each component of each illustrated apparatus is functionally conceptual, and does not necessarily need to be physically configured as illustrated. In other words, the specific form of distribution / integration of each device is not limited to that shown in the figure, and all or a part thereof may be functionally or physically distributed or arbitrarily distributed in arbitrary units according to various loads or usage conditions. It can be configured in an integrated manner (for example, the damage assumption unit 15a and the damage assumption update unit 15b immediately after a disaster are integrated). Further, all or any part of each processing function performed in each device may be realized by a CPU and a program analyzed and executed by the CPU, or may be realized as hardware by wired logic.

  The disaster damage estimation method described in this embodiment can be realized by executing a program prepared in advance on a computer such as a personal computer or a workstation. This program can be distributed via a network such as the Internet. The program can also be executed by being recorded on a computer-readable recording medium such as a hard disk, a flexible disk (FD), a CD-ROM, an MO, and a DVD and being read from the recording medium by the computer.

  As described above, the disaster damage assumption device, the disaster damage prediction method, and the disaster damage prediction program according to the present invention are useful when assuming damage to a predetermined structure due to a disaster that has occurred in a predetermined area, Suitable for making quick and accurate damage assumptions after a disaster.

It is a block diagram which shows the structure of the earthquake disaster assumption apparatus in a present Example. It is a figure for demonstrating a target area map and an earthquake history memory | storage part. It is a figure for demonstrating a damage rate function. It is a figure for demonstrating the area coefficient in a damage rate function. It is a figure for demonstrating the performance value in a damage rate function. It is a figure for demonstrating the performance value in a damage rate function. It is a figure for demonstrating a setting network structure memory | storage part. It is a figure for demonstrating the damage assumption part immediately after a disaster. It is a figure for demonstrating the conditional probability table used in the damage assumption part immediately after a disaster. It is a figure for demonstrating the estimation method by the Bayesian network used in the damage assumption part immediately after a disaster. It is a figure for demonstrating the estimation method by the Bayesian network used in the damage assumption part immediately after a disaster. It is a figure for demonstrating a damage assumption update part. It is a figure for demonstrating a damage assumption update part. It is a figure for demonstrating the process of the earthquake disaster assumption apparatus in a present Example. It is a figure for demonstrating another Example.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Earthquake disaster assumption apparatus 11 Input part 12 Output part 13 Input / output control I / F part 14 Storage part 14a Target area map / earthquake history storage part 14b Damage rate function storage part 14c Setting network structure storage part 15 Processing part 15a Damage immediately after disaster Assumption part 15b Damage assumption update part 15c Earthquake damage assumption map creation part

Claims (9)

A disaster damage estimation device that assumes damage to a predetermined structure due to a disaster that has occurred in a predetermined area,
Disaster history information storage means for storing disaster history information, which is history information of the disaster that occurred in the past in the predetermined area, for each mesh preset in the predetermined area;
From the performance value indicating the safety of the support installed in the predetermined area and the regional coefficient quantifying the regional characteristics in consideration of the ground conditions of the area where the support is installed, the disaster Damage rate function storage means for storing a damage rate function for calculating the damage occurrence rate of the support for each mesh based on the disaster information which is the information of
A Bayesian network expressing the occurrence of the disaster as a node without a parent node and expressing a causal relationship between a plurality of damage events including damage to the support and damage to the predetermined structure caused by the disaster with a conditional probability Setting network structure storage means for storing a setting network structure defined and set by
Estimating disaster information in the entire predetermined area based on immediately after disaster information, which is disaster information acquired immediately after the occurrence of the disaster, and the disaster history information for each mesh stored in the disaster history information storage means Then, the estimated disaster information in the entire predetermined area, the confirmed damage information immediately after the disaster that is the damage information determined in the entire predetermined area acquired together with the immediately following disaster information, and the damage rate function storage means In the predetermined structure immediately after the occurrence of the disaster, for each mesh preset in the predetermined area based on the damage rate function to be performed and the set network structure stored in the set network structure storage unit Immediately after the disaster is assumed to assume the damage incidence rate,
Based on updated disaster information that is updated disaster information after the occurrence of the disaster and damage information of the support determined in a part of the predetermined area after the occurrence of the disaster, the damage rate function Assuming the damage occurrence rate in the predetermined structure in the setting network structure stored in the setting network structure storage means, except for a part of the predetermined area, by updating the damage rate function stored in the storage means A damage assumption updating means for updating the mesh for each predetermined mesh in the predetermined area;
Disaster damage estimation device characterized by comprising   When the post-disaster confirmed damage information, which is damage information determined throughout the predetermined area after the occurrence of the disaster, is acquired, the damage assumption update means stores the setting network stored in the setting network structure storage means In addition to adding damage events corresponding to the post-disaster confirmed damage information to the structure, the assumption of damage occurrence rate in the predetermined structure based on the added setting network structure other than a part of the predetermined area, The disaster damage estimation apparatus according to claim 1, wherein the disaster damage estimation apparatus is updated for each mesh preset in the predetermined area.   A damage assumption map creation display that creates a damage assumption map in which the damage occurrence rate assumption result of the predetermined structure assumed by the damage assumption means immediately after the disaster is displayed together with a map of the predetermined area and displays it on a predetermined display unit The disaster damage estimation apparatus according to claim 2, further comprising means.   The damage assumption map creation / display means confirms a damage occurrence rate assumption update result of the predetermined structure in a part of the predetermined area other than a part of the predetermined area assumed by the damage assumption update means. A damage assumption map creating and displaying means for creating the displayed damage assumption map and displaying it on the predetermined display unit together with the damage information of the support and the confirmed damage information after the disaster is further provided. The disaster damage estimation apparatus according to claim 3.   The disaster is an earthquake disaster, the disaster information is seismic motion information composed of seismic intensity distribution, the confirmed damage information immediately after the disaster is power failure information, and the predetermined structure is the support The disaster damage assumption apparatus according to any one of claims 1 to 4.   The disaster is an earthquake disaster, the disaster information is seismic motion information composed of seismic intensity distribution, the confirmed damage information immediately after the disaster is power failure information, and the predetermined structure is adjacent to the support It is a building, The disaster damage assumption apparatus as described in any one of Claims 1-4 characterized by the above-mentioned.   The disaster is a typhoon disaster, the disaster information is weather information including wind speed information, the confirmed damage information immediately after the disaster is power failure information, and the predetermined structure is the support or the support It is a building adjacent to a thing, The disaster damage assumption apparatus as described in any one of Claims 1-4 characterized by the above-mentioned. A disaster damage estimation method that assumes damage to a predetermined structure due to a disaster occurring in a predetermined area,
Disaster history information storage step for storing disaster history information, which is history information of the disaster that occurred in the past in the predetermined area, in a first storage unit for each mesh preset in the predetermined area;
From the performance value indicating the safety of the support installed in the predetermined area and the regional coefficient quantifying the regional characteristics in consideration of the ground conditions of the area where the support is installed, the disaster A damage rate function storage step of storing a damage rate function for calculating a damage occurrence rate of the support for each mesh in the second storage unit based on the disaster information which is the information of;
A Bayesian network expressing the occurrence of the disaster as a node without a parent node and expressing a causal relationship between a plurality of damage events including damage to the support and damage to the predetermined structure caused by the disaster with a conditional probability A setting network structure storing step for storing the setting network structure defined and set by the third storage unit;
Estimating disaster information in the entire predetermined area based on immediately after disaster information acquired immediately after the occurrence of the disaster and the disaster history information for each mesh stored in the first storage unit And the second storage unit stores the estimated disaster information in the entire predetermined area, the confirmed damage information immediately after the disaster that is the damage information confirmed in the entire predetermined area acquired together with the immediately following disaster information, and the second storage unit. In the predetermined structure immediately after the occurrence of the disaster, for each mesh preset in the predetermined area, based on the damage rate function to be performed and the set network structure stored in the third storage unit Immediately after the disaster assumption step to estimate the damage occurrence rate,
Based on the updated disaster information that is updated disaster information after the occurrence of the disaster and the damage information of the support determined in a part of the predetermined area after the occurrence of the disaster, the second Assuming the damage occurrence rate in the predetermined structure in the setting network structure stored in the third storage unit, except for a part of the predetermined region, by updating the damage rate function stored in the storage unit A damage assumption update step that is performed by updating for each mesh preset in the predetermined area;
A disaster damage estimation method characterized by including A disaster damage prediction program for causing a computer to execute a disaster damage prediction method that assumes damage to a predetermined structure due to a disaster that has occurred in a predetermined area,
Disaster history information storage procedure for storing disaster history information, which is history information of the disaster that occurred in the past in the predetermined area, in the first storage unit for each mesh preset in the predetermined area;
From the performance value indicating the safety of the support installed in the predetermined area and the regional coefficient quantifying the regional characteristics in consideration of the ground conditions of the area where the support is installed, the disaster A damage rate function storage procedure for storing a damage rate function for calculating a damage occurrence rate of the support for each mesh in the second storage unit based on the disaster information which is the information of
A Bayesian network expressing the occurrence of the disaster as a node without a parent node and expressing a causal relationship between a plurality of damage events including damage to the support and damage to the predetermined structure caused by the disaster with a conditional probability A setting network structure storing procedure for storing the setting network structure defined and set by the third storage unit;
Estimating disaster information in the entire predetermined area based on immediately after disaster information acquired immediately after the occurrence of the disaster and the disaster history information for each mesh stored in the first storage unit And the second storage unit stores the estimated disaster information in the entire predetermined area, the confirmed damage information immediately after the disaster that is the damage information confirmed in the entire predetermined area acquired together with the immediately following disaster information, and the second storage unit. In the predetermined structure immediately after the occurrence of the disaster, for each mesh preset in the predetermined area, based on the damage rate function to be performed and the set network structure stored in the third storage unit Immediately after the disaster, assuming the rate of damage,
Based on the updated disaster information that is updated disaster information after the occurrence of the disaster and the damage information of the support determined in a part of the predetermined area after the occurrence of the disaster, the second Assuming the damage occurrence rate in the predetermined structure in the setting network structure stored in the third storage unit, except for a part of the predetermined region, by updating the damage rate function stored in the storage unit A damage assumption update procedure for updating for each mesh preset in the predetermined area;
Disaster damage forecasting program characterized by having a computer execute.

Images